This guest post comes from David Trombly, Data Management Engineer with the UC Davis Division of Utilities.
Building better campus data systems empowers users, operators, engineers, and sustainability professionals to do what they do best – save! Nearly all efforts to make the campus more sustainable rely on a foundation of accurate information about campus energy, water, waste, and chemical use. Developing more granular data streams opens the door to identify behaviors and processes that can be improved with reinforcing feedback loops. Managing these large data sets is a challenge. If we are to have a sustainable campus, we also need sustainable business processes for our data. Data must be properly collected, managed over time, run through quality checks, and periodically audited to keep it accurate and relevant.
In Utilities, we have been focusing on building more sustainable data management systems. Our metering and controls team has installed many new meters that measure campus buildings, wells, wastewater flow and treatment, and energy production and distribution. We are also working with Facilities Management to implement meter commissioning and calibration programs. Most of the real time meter data is now being automatically sent to a recently upgraded computer system which will allow operators, managers, researchers, students, and sustainability professionals to access the data from workstations and mobile devices as well as dashboards like the water dashboard and the Campus Energy Education Dashboard.
We are also building and automating database systems that track “meta data” – data around campus assets that provides context for real time data – such as process flow diagrams to help describe what exactly each meter is measuring. Much of this data will be combined with the real time data into one system and also displayed visually in collaboration with our GIS team, as seen here, where campus exterior lights are visualized for a section of main campus.
Here, campus exterior lights are visualized for a section of main campus.
This data will ultimately be combined into efficiency metrics and analytics which will more quickly reveal inefficiencies and trigger notifications of problems.
There are many cool things that can be done with big data for the small city that we call UC Davis. We are currently working on automating the recognition of building occupant behavior at South Entry Parking Structure. This relies on the fact that each device at a building has a unique electricity use signature. In the example below, the tall thin spikes are the elevator and the tall spikes following by slow ramp downs are car chargers. Once the algorithm is completed, we hope that it will help save energy, track occupant behavior, and automate building maintenance work orders based on changes to the signal.
South Entry Parking Structure, electrical power demand graph